mjs7231 | Hey guys, I'm having trouble creating a snap that uses Gtk. It seems everything is working fine, but when I try to run my Python script, I get "ValueError: Namespace Gtk not available" | 02:14 |
---|---|---|
mjs7231 | I'm probably missing a dependency, I tried gir1.2-gtk-3.0, gobject-introspection, and libgtk-3-0 -- nothing works. | 02:15 |
=== chihchun_afk is now known as chihchun | ||
Mikaela | Is there anything like apt:url for snaps? | 05:52 |
zyga | o/ | 09:09 |
=== JanC is now known as Guest47278 | ||
=== JanC_ is now known as JanC | ||
=== chihchun is now known as chihchun_afk | ||
=== chihchun_afk is now known as chihchun | ||
=== chihchun is now known as chihchun_afk | ||
=== chihchun_afk is now known as chihchun | ||
=== chihchun is now known as chihchun_afk | ||
=== chihchun_afk is now known as chihchun | ||
=== chihchun is now known as chihchun_afk | ||
=== chihchun_afk is now known as chihchun | ||
=== chihchun is now known as chihchun_afk | ||
sborovkov | Hi. Couple of questions about new security - is there a way now to say run root commands from snap? Particularly reboot. Also RPI specific - how would it be possible to be able to change config.txt which is a part of gadget snap from another snap? | 15:41 |
jdstrand | Trevinho: hi! did you happen to get hello-unity working? | 20:00 |
jdstrand | Trevinho: fyi, for the launcher API: https://github.com/ubuntu-core/snappy/pull/1173 | 20:01 |
Trevinho | jdstrand: yeah... | 20:01 |
Trevinho | jdstrand: I mean, I got the hello unity thing working | 20:01 |
Trevinho | jdstrand: I wanted to push it, but I've not done it yet | 20:01 |
Trevinho | jdstrand: I'll do that in a bit | 20:01 |
jdstrand | Trevinho: thanks! | 20:02 |
Trevinho | jdstrand: do you happen to know why gtk apps get the anonymous path? | 20:04 |
Trevinho | jdstrand: it should be appid... | 20:04 |
Trevinho | That's something we also noticed at the desktop sprint | 20:04 |
jdstrand | no idea | 20:04 |
Trevinho | jdstrand: the only thing I don't get working in hello unity is the indicator icon... I didn't debug that much yet, though... But I believe it's libappindicator not being able to locate the icon (although I've all the caches set) | 20:07 |
jdstrand | Trevinho: intersting. If you want to commit what you have, I can verify what you have so far | 20:18 |
Trevinho | jdstrand: yeah, give me 10 minutes, since I've done that in a VM which is in a different machine from the one I'm on right now :) | 20:19 |
jdstrand | np, thanks again | 20:21 |
Kristbaum | Hello, is this the right place if I have questions about snappy? | 20:21 |
jdstrand | Kristbaum: yes | 20:22 |
Kristbaum | Okay thanks ;) I wanted to know what the "home" slot in "snappy interfaces" is about. I coudn't find anything elsewhere.. | 20:24 |
jdstrand | Kristbaum: the docs are being worked on and are a bit behind atm on the website. You can see this though: https://github.com/ubuntu-core/snappy/blob/master/docs/interfaces.md | 20:25 |
Kristbaum | Ah thanks, so if I enable the "home" slot in the .yaml file, what happens if a user tries to install it? | 20:28 |
jdstrand | the interface is not (currently) autoconnected, so the access is denied | 20:29 |
Kristbaum | Can a User enable it? | 20:29 |
Kristbaum | with snap connect? | 20:30 |
jdstrand | the user would then need to do something like: snap connect <you snap name>:home ubuntu-core:home | 20:30 |
Kristbaum | But support for autoconnect is planned? | 20:31 |
jdstrand | autoconnect is there now | 20:31 |
jdstrand | it is explicitly set to not autoconnect | 20:32 |
jdstrand | aiui, the gadget snap would make that auto-connectable | 20:32 |
jdstrand | also, future snap assertions might make then autoconnect depending on how the store is set up (that is planned) | 20:32 |
jdstrand | them* | 20:33 |
jdstrand | also, tooling might improve to assist with connecting | 20:33 |
zyga | jdstrand: o/ | 20:34 |
Kristbaum | Ah, so maybe in the future, there may be a Dialogue where the User decides if he want's to enable access tho his/her homefolder? | 20:34 |
zyga | jdstrand: I was off today, sorry for not realizing this earlier | 20:34 |
jdstrand | Kristbaum: it's conceivable | 20:34 |
jdstrand | hey zyga :) | 20:34 |
jdstrand | zyga: if you're off, why are you here now? | 20:35 |
ogra_ | jetlaggers paradise ;) | 20:35 |
zyga | jdstrand: just talking to slangasek about a missed meeting | 20:35 |
jdstrand | Kristbaum: you might be interested in https://insights.ubuntu.com/2016/05/04/security-confinement-in-ubuntu-core/ | 20:36 |
zyga | jdstrand: ohh, published now! | 20:36 |
zyga | jdstrand: cool | 20:36 |
jdstrand | zyga: that's niemeyer's blog post from a little while ago | 20:36 |
jdstrand | not the whitepaper | 20:36 |
zyga | ahh, too bad :) | 20:36 |
zyga | I look forward for the whitepaper | 20:37 |
jdstrand | Kristbaum: there is also http://www.zygoon.pl/2016/04/snappy-snapcraft-and-interfaces.html | 20:37 |
jdstrand | zyga: it will be nice to have up. I think it is in the community team's court at this point | 20:38 |
Kristbaum | jdstrand: Thanks for this ;) How exactly are the snaps on desktop Ubuntu updated at the moment? Do I have to hit refresh on ervery app? | 20:38 |
ogra_ | there used to be an autorefresh feature, that was disabled and i currently re-worked | 20:40 |
ogra_ | it will return eventually | 20:40 |
ogra_ | (currently you have to use snap refresh) | 20:41 |
Kristbaum | Interesting. Do you know anything about how source code will be managed in the snappy store? Are there considerations on this? (In the sense of "apt source") | 20:44 |
ogra_ | no, snaps are purely binary | 20:47 |
ogra_ | you woud just host your code on github or wherever and aadd a snapcraft.yaml to the tree | 20:48 |
Kristbaum | Isn't this a Problem when you want a trusted, reproducable build? | 20:49 |
ogra_ | why would you ? | 20:49 |
ogra_ | the esssence of a snap is that it can be competely untrusted thanks to the confinement and the entry checks in the store | 20:50 |
Kristbaum | If, for example, I use a FIrefox snap to to 90% of my work, how do I know that the snap is really only the sourcecode from Mozilla, not anything else. | 20:50 |
Kristbaum | ? | 20:50 |
ogra_ | you cant ... | 20:51 |
ogra_ | i mean ... iit is likely that it is unmodified if mozilla uploaded it to the store ... but there is no guarantee | 20:51 |
Kristbaum | Yeah, but isn't this a problem, or am I seeing this wrong? | 20:52 |
ogra_ | and effectively it does not matter ... since you have to decide what you allow to that snap | 20:52 |
Kristbaum | But I do need to trust Firefox, because my work is in it, not in the Rest of the System. And if i can't check if it's really Firefox, how does this help me? | 20:54 |
ogra_ | how do you know it today ? | 20:55 |
jdstrand | Kristbaum: that gets into the assertions stuff I referenced earlier | 20:57 |
Kristbaum | I guess I trust the Debian/Ubuntu Developers that took the source and put it in the repo, or I type apt source and check it myself. | 20:57 |
jdstrand | Kristbaum: the idea is that you'll be able to be confident that a snap comes from a particular publisher (eg mozilla). assertions have crypto the ensures that and that crypto could be used to verify your installed snap | 20:58 |
jdstrand | Kristbaum: that does not guaratee that the publisher built the snap from their source trees | 20:58 |
Trevinho | jdstrand: sorry for the delay, i'm pushing the thing soon... but i've just noticed that I've lost some of the things I was working on (damn me and staing on tmp), so... let me check i can get them back | 20:58 |
jdstrand | Kristbaum: Launchpad will grow functionality to build snaps, similar to ppas today, that can be triggered off source uploads. some of that is already in place | 20:59 |
ogra_ | it already has it | 20:59 |
jdstrand | yes, but not autouploads to the store | 21:00 |
ogra_ | as a button for a tree though ... | 21:00 |
Kristbaum | Oke, this isn't bad, so I can at least trust, it's from Mozilla. And will these Launchpad snaps be marked, so I know I can check the source if I want, and they directly are build from this source? | 21:01 |
Kristbaum | jdstrand: ^ | 21:01 |
jdstrand | I'm not up on those details. Immediately you could download the snap from Launchpad, generate a hash of the file and compare that to what is on your system | 21:02 |
jdstrand | I know source builds are something that is desired; I just don't know how they tie into assertions | 21:03 |
Kristbaum | Ok, but it isn't on the roadmap yet? You would get a lot of "free-software-street-cred" if you implement this, I would presume. :D | 21:05 |
jdstrand | it is on the roadmap, I just don't know the plans | 21:06 |
jdstrand | Kristbaum: you might ask on the list-- you'll get more authoritative answers :) | 21:07 |
jdstrand | I think a lot of people are off today | 21:07 |
ogra_ | at least we pretend to :) | 21:08 |
Kristbaum | Thanks, you helped me a lot ;) ! Very nice talking to you, I'll be off then, good night! | 21:09 |
qengho | jdstrand: it would be kind of cool to have reproducible builds and ship a snapcraft.yaml if it came from that. | 21:24 |
qengho | Ask user at upload time. | 21:25 |
qengho | Ask publisher, that is. | 21:25 |
jdstrand | I would love reproducible builds | 21:26 |
niemeyer | jdstrand, zyga: We need to work on a disc-drive interface of sorts, soonish.. | 21:33 |
jdstrand | niemeyer: I have no details on that. can you file a bug? | 21:43 |
jdstrand | (with the snapd-interface tag of course :) | 21:43 |
=== blr_ is now known as blr | ||
niemeyer | jdstrand: Will do, although I don't have many details either ;) | 22:43 |
mhall119 | sergiusens: where can I find the snapcraft.yaml for your telegram snap? | 22:44 |
niemeyer | jdstrand: The idea is just enabling applications that want /dev/cdrom etc | 22:44 |
niemeyer | mhall119: Under his GH username | 22:44 |
niemeyer | sergiusens/telegram-snap | 22:45 |
mhall119 | thanks niemeyer | 22:51 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!